Page 159 - DCAP606_BUSINESS_INTELLIGENCE

Business Intelligence

Notes
Example: Classes can be defined to represent the likelihood that a customer defaults on
a loan (Yes/No).
A common approach for classifiers is to use decisions trees to partition and segment records.
New records can be classified by traversing the tree from the origin through branches and
nodes, to a leaf representing a class.

Figure 11.3: Classification Example

Source: http://www.siggraph.org/education/materials/HyperVis/applicat/data_mining/images/
tree.gif
Regression

The dependency of one or more independent predictor variables on a single response variable
is modelled using regression.

Task Compare and contrast the clustering and classification analysis.

11.3 Data Mining Uses

Data mining is utilised for a variety of reasons in both the personal and public parts. Industries
such as banking, insurance, medicine, and retailing commonly use data mining to decrease
charges, enhance research, and increase sales.

Example: The protection and banking industries use data mining applications to notice
deception and aid in risk evaluation.
Using customer data assembled over several years, businesses can evolve forms that predict if a
customer is a good credit risk, or if whether misfortune claim may be fraudulent and should be
investigated more neatly. The medical community sometimes utilises data mining to help
forecast the effectiveness of a procedure or surgery.
Pharmaceutical companies use data mining of chemical compounds and genetic material to help
direct study on new treatments for infections. Retailers can use data assembled through affinity
programs (e.g., shoppers’ club cards, common flyer points, contests) to consider the effectiveness

154 LOVELY PROFESSIONAL UNIVERSITY